ScreamingBox Technology & Business Rundown

Dave Erickson 0:00
Have a hard time looking backwards, reluctant to try to predict the future? On this ScreamingBox podcast, we're going to look both backwards and forwards at the Business and Technology events of the last year and the coming year. Please like our podcast and subscribe to our channel to get notified when the next podcast is released.

Dave Erickson 0:41
It's hard to believe, but this is our 62nd monthly podcast, and everyone wants to know why is this a thing. Welcome to the screambox technology and business rundown podcast. In this podcast, I, Dave Erickson and my honorable co-host Botond Seres, are going to look back at 2025 and then look forward to what 2026 might bring. No guests for this podcast, just the two of us looking back and looking forward at the same time. So Botond what did you think about 2025?

Botond Seres 1:11
Well, I think it's been, in general, a very exciting year for tech and new software, new hardware, new AIs. Personally, for me, it's been an extremely challenging year, so I wonder how has it been for you, Dave?

Dave Erickson 1:33
It's been a very busy year, and I like the idea that we were able to make a lot of progress in an industry that's changing really fast. AI has really kind of brought in a lot of different changes, and a lot of our work has kind of changed a little bit. We did more kind of healthcare AI type of projects helping companies try to figure out how to go through data and make sense of it and then apply it to their business challenges. But also it's a very volatile year. A lot of developers are moving around. Developers are also looking for work more than they were, say, a year ago, because projects have shifted and moved around. But also the needs are different. I talk to a lot of developers, and the thing that is exciting to them is that they have to learn new stuff that's really good. But obviously from a business standpoint, it was a challenging year for a lot of people. I don't like to say it was challenging for us, it was changing, and things change, and that's okay. We're pretty flexible and that’s good, but I'm, I'm kind of excited about all the different technologies and where development is going. What about you in 2025 What did you learn that was new?

Botond Seres 3:01
Oh, well, I mean, I learned that AI video generation tools are incredible at this point. It's, it's no longer a matter of looking at the fingers in the generated image or the teeth. They are becoming more and more realistic, which is quite concerning. I also learned that in coding, there certainly are uses for AI, which I didn't think was the case in 2024 since, let's say the output was quite inconsistent, but now it's getting better and better. Also , it's incredibly interesting to me how this appears to be the year when we finally started to implement safeguards for a lot of things regarding this new technology, and I'm really excited about how that will develop and how we will get more and more RS that will need to be handled in a very careful way, also regarding protected classes of information, like healthcare information. As you said, it's, it's one of the things that was not really a concern a couple years ago, but now it's, it appears to be in the spotlight, but I'm sure you have a lot of things you learned this year, Dave.

Dave Erickson 4:41
I think the most interesting thing that happened in 2025 from a technology standpoint, was actually when we were discussing with our, one of our last guests, Ari Block, about the math of LLMs and AI and that, how it's progressed and where it's going. But I could definitely see AI tools in the beginning of the year and AI tools at the end of the year, and in one year, such growth has happened in those tools, right? And I've talked to a bunch of developers who are learning a lot of these AI coding tools. You know, an individual developer right now is able to get maybe a 20-30% productivity boost by using these tools, right, and 20-30% is a decent amount of productivity growth from a tool, and can definitely help developers, you know, reduce the amount of time needed to do certain things. Obviously, it's, it's more beneficial in simple things. So if you need a login screen, and, you know, you have one of the AI tools code you up a basic login screen that's relatively easy, and that'll save you a bunch of time, in the sense that it's a small amount of time, and it'll save you a lot of that; a complex task, you know, it might help you 10% right, because you're gonna have to go through it and check everything. And, you know, it may not be good enough, or it may not be complex, but may give you a good base to start coding from right I've also talked to some other developers who work in a team, and they find that if they use the AI coding tools as a team, they actually can get more like 30 - 40% productivity bump. But that has to come from a lot of inner team coordination and communication of using the tools correctly. And so, you know, individual developers will kind of see one result. A team of developers who are coordinated will see a different result. But I think as far as development is concerned, it's going to mean more productivity, and I think it's going to segregate out developers. There's going to be those developers who know AI tools and are good with AI tools and get that productivity bump. And then there'll be developers who are, who are not interested, and they just want to code like they normally code, and they may get their work or certain types of work, but they'll be excluded from a lot of other work. And, you know, I think that's just kind of the direction it's going. What do you think?

Botond Seres 7:35
Well, I do think that there is already a segregation, but not necessarily with AI at this point. It's more a segregation of developers who can properly use tools like IntelliSense or Resharper to not only double check their work, but also to even generate code much, much quicker than traditionally, just writing it on the keyboard. And I feel like that's the part that is getting turned up to 10 with generative AI, because at this point, we cannot only use these rule based very strict systems that evaluate our work continuously and help us write better code, but there are some parts that we don't even have to write anymore. And at the end of it all, it is really important that we learn to use these static analysis tools to verify that the generated code is actually good or not, which is, let's face it, more often than not, is not great, but with the combination of these two technologies, it is now remarkably easy to write code that was just deemed too difficult or too not really too difficult, but too tedious. Like, perfect example of this is Auto Mapper, which, all it does, you have one object, you have another object, both of them have some properties, and then you want to assign the values from one to the other. And personally, I'm really glad that we can put this behind us, because it has a billion issues in production. There's always some edge case where it doesn't work, and with, with the help of AI, we can finally just generate those assignments, and then, with the use of static analysis tools, we can actually verify them. And I think that is absolutely amazing. And I do think that's the kind of difference that we are going to see in how people code, because previously, it was quite accepted to just not do these things and take lots and lots of shortcuts, but now that we can generate just this huge mass of codes in seconds or minutes, personally, I believe there is no more reason to take these shortcuts anymore. Additionally, you brought up Ari, and I think he has an incredibly exciting outlook on AI in general. And one thing that stood out to me during our podcast with him, is that I read a science fiction book which seems to be based on the premise of the thing that he's actually working on at this moment.

Dave Erickson 10:32
And what was the name of the book?

Botond Seres 10:35
Oh, it's Fly Bots fly, as in an actual fly. So the premise is that eventually we're going to get to a technology where robotics is so miniature that we can get robotic flies. And additionally to this, the main premise also includes that we are not going to have these massive AI clusters. Instead, we are going to rely on these tiny, little AI brains that will be able to work in sort of a hive mind. And he seems to be working on that exact thing, which is amazing.

Dave Erickson 11:14
I think that that's very realistic. And I think that that is an edge case tha’s definitely going to come in. I mean, robotics is advancing very quickly right now, and you know, the next stage of that is basically combining robot robots, independent robots, right with AI, and then they're operating kind of autonomously. I think we're still a ways away from having that happen in a very refined and secure way, but a ways away maybe five years, maybe 10 years kind of thing. And at that point, people are going to have an issue, because robots will be able to do things that people can do, and so that's going to create its own issues. But also the miniaturization of robots and the form of robots is going to allow for lots of different things to happen that will be very positive as well, right? And you know, computers are not limited in a way that humans are. So the possibility of a hive mind is very possible, and it's kind of done now, right? If you look at how AI agents are currently working, a lot of them are communicating with each other, quote, like a hive mind. They're, they're coordinating, they're, they're moving in a certain direction. If you architect the AI agent structures in a way, then a bunch of different things, a bunch of different agents act as one, right? And that's kind of a hive mind. It has its own technology issues, and one of those is memory, because they all have to work from kind of the same memory. And when you write these agents, if you're not careful, you're going to have different things put in different places, and so they can't access the same memory. So some of them forget things, and some of them know things that the others don't and you know, but all that stuff will get worked out. It's just a matter of time. So, yeah,

Botond Seres 13:27
What did you think about Carley Martha? Personally, I think she has a very pragmatic approach to AI. She's actually the only person I know that has used agents successfully, and I think it's really great how she's using them as intended for very simple and very repetitive tasks.

Dave Erickson 13:53
Yeah, I agree. And she was one of the more interesting guests that, that we had, and I really think that the, the concept of agents, what's, what's positive about it right now is that on a small enough level, it's something that can be used by most businesses relatively quickly and easily, right? And I think that the big, the big help of AI right now is to make it so that small businesses can operate at a much larger level without having to have all that overhead, right? And I've been using AI agents and AI in general to help other businesses figure out things, and a lot of it is surrounding business challenges, and I've written out AI agents who help the clients go through a series of interviews and workshops that allow them to actually say what they want in a detailed enough way where I can create an SOW from it and give it to developers to do work. Right? And one of the biggest issues that we've had as a company is the client. They kind of know what they want, but they can't, you know, and they may even have an SOW, but when you look at the SOW and then we hear what they want, it seems like they're asking for two different things. And so having an AI agent, you know, run through a bunch of questions with the client and then take that information and format it into an actual SOW that is what they want. That's actually been a very powerful tool. And a year ago, it would have been very hard to do that, right?

Botond Seres 15:58
Yes, I do know that working on SOWs in general, is thankless and very difficult work. So it's really great to know that this new tech can help with that in some ways.

Dave Erickson 16:13
Yeah, it, it took a while to figure out. What is the interview process, what are the questions you're asking? What is the workshop? Because you have to kind of give information as well. These are all the different options. What is an option that fits your business better? What is your business challenge? I guess the advantage I have is I've done so many different types of businesses and worked with so many different types of clients that I have a feel for where the questions need to lead to, right? And you have to do it kind of in a way that is a little respectful, because the client comes to you thinking they know what they want. And you know, if you've done enough of these projects and talks with clients, yeah, some of them actually know exactly what they want, and they're able to communicate it well, and that's great.. But there's also lines, yeah, but there's also a lot of them that they don't really know what they want, or they have a team, and they kind of want different things, and so it's really important to kind of put something together that they can work with as a tool. I mean, I have an example, a client came to me. They have a website, they have a business, but they wanted to make their website do something, because it was, at this point, just kind of a marketing website, and they wanted it to develop their business. And so I went through a whole workshop with them, gathering up information based on, kind of, a format of questions I had developed for my AI interrogations. And then I was able to take all that data that they gave me and feed it into the AI and come up with not just an SOW for us to build a website, but a whole brand marketing book for them, so that they could generate a lot of the content for the website, for their social media, for everything surrounding the business while we build the website. Right? So it provided an additional service, and it didn't cost me anything extra to provide that service. And it got me the SOW, but it got them the things they need to feed me, everything for the website. So they know what photos to take. They know what their taglines are. They're going to have to take this information and use, you know, develop a logo from it. But beyond that, they're really set up to do business, and we're able to basically build a website quickly, and it won't, you know, break the bank for them, and it won't cost us enough so we can actually make a profit with it.

Botond Seres 18:55
That's great. So they don't have to contact an external firm to just get that SOW put together. It's an automated service now, which is another guest we had on the podcast is Nono Bokete. I hope I'm saying that right, and I really liked how she had a remarkably realistic approach to AI, I mean, she seems to be aware of both the risks and the benefits, and appears to have a quite balanced opinion of them. Personally, I tend to lean towards worrying about the risks rather than being happy about the benefits. Some of the benefits seem to be quite small compared to the risks and I, I do know that you have quite an optimistic opinion. Dave, on, on, on, AI in general. And that is, if you don't mind, it could be our next topic like, what do you think about some of the risks and benefits of AI in general?

Dave Erickson 20:14
Well, I think you know. So the way humans normally think, if they can invent something that saves them work and means that they don't have to work so hard to get something that's positive. And in the sense of AI, that is a reality. But I think that absolutely people are not quite understanding the scale at which it will be able to save work. And when that is combined, not just with what AI can do, but also when you give it a physical component, by putting it in robots, right, it's going to be able to do a lot of what people do now and corporations who basically, now at this point, run the world, are in it for the money are in it for profit, of the for the people who own the company, not for the people who work in the company. And if they have to make a decision, hey, I can save money by getting robots and AI and not have people working, they'll do that. And I think that at some point there's going to be a situation where there's people who are just not able to work because the jobs aren't available. They've been taken up by AI and robots. And that's okay, if the world has a plan for that, but if the world doesn't have a plan for that, then you're just going to have a bunch of people who have no work and they need a place to live and they need food to eat, and society is going to have to figure that one out right. Now, in the Future, utopia of the future, if AI and robots are doing everything and people don't have to work well, if they're produ, if the robots are producing all the food and they're giving the food away for free, and they're building houses and they're giving the houses away for free, I guess it will work, but our societies are not set up for that; that's not the way it works now. So huge changes will have to be made to accommodate that, unless the world decides now, hey, let's limit AI so that there are jobs. And you know, so the world's got to say maybe to the companies, hey, we know you would like to increase your margin by replacing a whole bunch of people with AI and robots, but we're not going to let you do that. You're just going to have to be inefficient, because we need people to work so they can buy your products, and they can buy food, and they can do all this other stuff. There is a problem brewing, and I don't hear a lot about solutions yet.

Botond Seres 23:00
Yeah, I'm not sure that the restriction path is going to work out great. I mean, I think pretty much everyone is familiar at this point with some of these future utopias that we might be mentioning. And I think even if not future utopias, but visions of the future. It's, it's quite interesting to me that lots and lots of these visions were developed, maybe 50 - 60, maybe even more, decades ago. And one of the things that they seem to be, seem to have in common is complete lack of AI. Like even in Star Trek, of course, you have the ship's computer, but it's not really a true AI. It's more like this sort of rules based thing that is very it's an extremely advanced theory, but it's not GPT, yeah, and or in Dune, they have this whole whole thing. I'm not very familiar with the methods of that world, but it appears that there was a revolution, I think, against AI taking all of the jobs, like, literally, all of them. And that's great. I mean, that's very interesting that at a point in time where we didn't even have AI many, many people were able to look through the veil of time and see that at some point this is going to be an issue in our future, and it may be a trigger to to societal change in some ways, like the Industrial Revolution was. I hope it's a change for the better. At this point, I think we need to, have to rethink loads and lots of things. Like, it's just not it's. Society is not going to work the way it is. I mean, we can try giving people more and more meaningless jobs, but even at this point in 2025 there is a massive amount of meaningless jobs, and we can, of course, provide more, but when do we stop? What's the point where we say that let's just have people have their freedom. Instead of doing a nine to five, let's just have the robots take care of everything and let the humans live. Is that the future that, that is even in the cards at the moment? And I think it might be, but I don't think it's in the near future, and I don't think that that future will come without a cost unfortunately.

Dave Erickson 25:51
Yeah, I agree, and I don't think the changes are going to happen. I mean, 20 years, 30 years from now, I think we're going to see a lot of those changes in the situation having to be something very different. I think in the next 5 years, 10 years, you're going to see the, some negative effects of it as well as positive effects of it. But societal pressure will not be that great in that short a period yet. But I go back to something, I've actually met some very wealthy people in my life, and one of the things that I have found with these people is that they're bored. When you have all the money you can possibly spend in the world, and you have bought all the things that you can buy, and gone on all the vacations you can go, you get to a point where you're bored. Nothing takes an effort when you're that wealthy and I've had these conversations where they are bored and they're afraid that they're going to be bored the rest of their life, because they don't have to make any effort to make something happen, they just buy it right. Every problem is solved with some money, and there's very little challenge. And after they've bought everything and gone everywhere and done everything. What else is there to do? Which is why you still see people who are wealthy beyond belief starting new businesses and trying new things. They're not doing it because they have to. They're doing it because they're bored and they want to do something. So that's the challenge that is going to face humanity if AI and robots do everything, people will be bored in the sense that what is their challenge. And you can be bored in a really boring job, and there are plenty of boring jobs out there, but you still kind of have a purpose at least, and you have to do it because you have to put food on the table and pay rent and all that, there is a challenge of just living. But if you didn't have that challenge of just living, if where you live and the food you eat is just given to you, there's kind of a lot of people will have emotional issues with that, right? That's the solution, I'm afraid of, is that if the robots take over everything, you're going to have, you know, 8 billion people who are pretty bored, and yes, they'll have the freedom to do anything, but, you know, once they've done it, what? Where's the challenge, right? It that it's why I keep doing businesses and why I keep trying new things, because that's what's interesting for me, right. And I think that's an issue with other people. I think that's a bigger, kind of, challenge that humanity will have is, how do you have robots and AI do everything and keep the population happy?

Botond Seres 29:04
Yeah, you're absolutely right. That's why I have as many hobbies as I can financially afford. But if I could just afford any and all hobbies, they, they wouldn't work anymore. Yep, that's why do I keep preparing my ancient sports car because I don't have the money to have someone else do it right. That's the main reason. If I just have a black AmEx and just say to any mechanic pay, just make this good. It's, It's too easy. But I do think that there are other motivations besides money, and I do have a feeling that I read somewhere that you can only truly know a man once you've known him through a long stretch of unemployment, I'm paraphrasing this extremely heavily, so please bear with me, but, but that's the idea. So when someone lacks purpose, what do they do? Do they keep working on themselves? Do they try picking up new skills, or do they just sit on the couch and watch any streaming service all day? And I don't know if this. it's different types of people, or if it's just different phases of life, but certainly there, there are examples of both.

Dave Erickson 30:37
Well, you see a kind of almost future, predictor of it a little bit at least. I see it and I see it in kids. There are kids who don't have much family structure, and they go to school, and after school, they come home and they sit on the couch with their iPad or their phone, and they look at that all day. The rest of the day, they don't do anything that they're literally trapped in that device. Whatever they're doing, playing games, chatting with friends, doing they can spend five hours, six hours staring at their phone on the couch after school, and they're happy doing that, maybe, right?

Botond Seres 31:27
at that point. Is that an escape, or is that just their world?

Dave Erickson 31:31
So that's kind of a future predictor in a sense, of what would happen if you got to have, if you had nothing to do, right? And it's going to get kind of worse with AI, because a lot of the content will be generated by AI. So AI is not great at creating, it's really good at taking what's already been created and reassembling it. So I think even entertainment generated by AI is going to stagnate after a while.

Botond Seres 32:01
Maybe that's the last bastion of humanity's entertainment. That's where all of us are going to have jobs once the robots take over. Yeah, everyone is going to be an actor in their own series, which is not that far off from where we are with YouTube and bloggers, or Tiktok even which, which has expanded the number of content creators even more. So I can see the direction we might be added in. It's quite, quite interesting.

Dave Erickson 32:38
Yeah, this concept of content creator as, kind of, a job. I mean, there's always been content creators, but it's always been so difficult to generate entertainment content that it was managed by large studios, large businesses and other things. And now a content creator can be anybody with a cell phone standing somewhere, and it's it. I think the problem it generates is you have so much content being generated that it's so hard to find any content that's actually worth watching, right?

Botond Seres 33:13
I wholeheartedly disagree. There's so much content worth watching and listening to.

Dave Erickson 33:20
So, I mean, there's lots of good content; finding it and consuming it is more of an issue. And even then, the content creators. You know, I follow several content creators who do things, and after watching them for six months or a year. A, you burn out on them, and B, they burn out on themselves, because they can only come up with so much. After six months of publishing every day or every other day a piece of content, they've kind of run out of ideas that's within their scope of life. And so they're kind of repeating each other, themselves after a while, and so, you know, I watch a content creator for maybe three or four months, and then I don't need to see them anymore, you know. So that's an issue.

Botond Seres 34:13
I think that's totally normal. But as much as you don't want to watch them anymore, there is always someone else who hasn't found them yet. So it's, that content creator is not really yours. It's just yours for the time being. Yes, or you are theirs as a viewer. It depends on where you look at it. But yeah, I really like how you brought up studios in general, or just big, wild groups of people working on content, and there seems to be a push to creating quality content now, because even before AI really became useful as a writing tool it has already become overused to the point that we are going back to the roots and hiring actual writers for big ticket projects, at least, that's what to be observing.

Dave Erickson 35:14
I think so, because it's, it's about creativity. So AI is really not good to create, creating people think it creates, but it's really just reorganizing old data. People like creating. People also like creating in groups, right? These are human activities, and when you take that away by using too much AI and other things, it's, it doesn't lead to happiness for people. And I think that's why, you know, I said earlier on that, you know, governments or people or agencies may have to say, Hey, don't use AI in this situation, use people. But I think it's also maybe people will say that, or people will say, Hey, I would rather create this with a group of people. It is not the most cost efficient, it is not the most profitable, but we enjoy doing that, so we're going to do it this way even though we have this alternative out there, we'd rather do it this way. And I think you'll see a lot more people push on that as AI becomes more invasive?

Botond Seres 36:22
Oh, absolutely. I think we can even look at this issue through the lens of the volume. I don't know if you're familiar, no, it's a thing in Hollywood. It's basically just a massive array of all that screens in which you have, I don't know how many square feet, but maybe like 10 square meters, like 10 by 10, it's a really small area, and there is, there are so many scenes that are being shot there, because you have the real background. You, you have the, even if you just have, like, some preliminary animation on there that they can touch up later, the actors know where they are supposed to be, what they are supposed to be seeing in a well, maybe not fully, but heavily cgi movie. And when the reactor actually goes out of their way and gets a proper location, gets a proper set where people can walk more than like 20 feet in the direction without bumping into a screen. There is just a widely positive response from the public. They may not understand what is different about that movie or series, but they can definitely feel it.

Dave Erickson 37:46
Humanity isn't lost in AI yet, there are possible directions. I think AI is going to bring some very interesting benefits and make things possible. And I think that's going to be especially felt in, you know, the transition from people working in a job to owning their own business, because they have no other choice, and AI will allow them to do that, in a sense. But there's also going to be the negative consequences of, well, there are people who don't have jobs because AI took it over and, you know, and it's a pendulum. It's gonna be swinging back and forth for quite a ways, right? I see it in marketing right now, earlier in the year, it was very hard to get marketing jobs because everybody was using ChatGPT to generate marketing content. And now it's a little bit easier to get marketing jobs because everybody realizes, okay, they generated a lot of AI marketing content, but it was kind of boring, and they need some creative ideas. They need something that's different and exciting. So they're swinging back to hiring marketing people again, right? So you're going to see this in all kinds of fields. This, this kind of moving back and forth, figuring out where AI fits in and where it doesn't.

Botond Seres 39:02
Oh yes, I think that's the key finding where AI fits in in the workflow. So it is great, it is amazing for certain things and absolutely terrible for other things. But en masse, we don't know what things it is good at and what things it, it's terrible at.That's the main, the main issue right now.

Dave Erickson 39:33
Well, today we've summed up 2025 and we forecast it out for 2026.

Botond Seres 39:39
So that means we are at the end of the episode today. But before we go, we want you to think about this important question.

Dave Erickson 39:48
How are you going to make an impact next year?

Botond Seres 39:52
For our listeners, please subscribe and click the notification to join us for our next ScreamingBox technology and business rundown podcast, and until then, have a great new year.

Dave Erickson 40:05
Thank you very much for taking this journey with us. Join us for our next exciting exploration of technology and business in the first week of every month. Please help us by subscribing, liking and following us on whichever platform you're listening to or watching us on. We hope you enjoyed this podcast, and please let us know any subjects or topics you would like us to discuss in our next podcast by leaving a message for us in the comment sections or sending us a Twitter DM till next month. Please stay happy and healthy.